est d'un naturel vif et emporté, et il est doué d'une intelligence faible et bornée. (p. 370.)... De là, il fit sept cents li au nord-ouest, et arriva au royaume de 'O-nant'o-pou-lo (Ananda-poura) (ce pays dépend du royaume de Málva). De là, il fit cinq cents li au nord-ouest, et arriva au royaume de Sou-la-tch'a (Sourâchtra). Ce royaume a environ quatre mille li de tour; la circonférence de la capitale est de trente li. Du côté de l'ouest, ce royaume touche à la rivière Mahi. Sa population est trés-nombreuse, et toutes les familles vivent dans l' abondance. . . . Comme ce royaume se trouve sur le chemin de la mer occidentale, touts les habitants en retirent de grands avantages, et font du commerce leur principale occupation. (Il est soumis au royaume de Falápi,-Siyuki x. 18). De là, il fit dix-huit cent li au nord-est, et arriva au royaume de Kiu-tche-lo (Gourdjara). Ce royaume a cinq mille li de tour. La capitale, appelée Pi-lo-mo-lo (Vîramâla ?) a trente li de circonférence. (Le roi est de la caste des Kchatriyas: Siyuki, xi., 18). Ensuite, il fit deux mille huit cents li au sud-est, et arriva au royaume de Ou-che-yen-na (Oudjayana). (Le roi est de la race des Brahmanes: Siyuki, xi. 18.) De là, il fit neuf cents li au nord-est, et arriva au royaume de Mo-hi-chi-fa-lo-pou-lo (Mahêçvarapoura). (Le roi descend d'une famille de Brahmanes.) De là, tournant à l'ouest, il revint au royaume de Sou-la-tch'a.'1

As a conclusion to this series of extracts and to put my readers in possession of the statements of Al Birúní in all their integrity, I append the French translation of all his observations on the Gupta era, inserting likewise the original Arabic of the most important passage.

'On emploie ordinairement les ères de Sri-Harcha, de Vikramaditya, de Saca, de Ballaba, et des Gouptas. . . . L'ère de Saca, nommée par les Indiens Saca-kâla,' est postérieure à celle de Vikramaditya, de 135 ans. Saca est le nom d'un prince qui a régné sur les contrées situécs entre l'Indus et la mer. Sa résidence était placée au centre de l'empire, dans la contrée nommée Aryavartha . . . . Vikramaditya marcha

1 [Since these notes have been set up in type, I have had an opportunity of perusing M. Julien's new publication, entitled, Mémoires sur les Contrées Occidentales, traduits du Sanskrit en Chinois, en l' an 648; par Hiouen-Thsang,' (vol. i. Paris, 1857). This work, though entering into more voluminous detail regarding the question of the Buddhist faith and the credulity of its votaries, contributes far less ample information in its historical references, than the previous publication from which my extracts are taken.

Huen-Thsang notices several of Asoka's and other Láts, sur le sommet de laquelle on a sculpté l'image d'un lion. Sur les côtés, on a gravé l'histoire du Nirvâna (de Krakoutchtchanda). Cette colonne a été construit par le roi Açôka' (p. 315, 316, 346; Benares, 354; Sarnáth, 355; Vaiçali, 387: etc.); and again, near Kousinagara, 'on a élevé, en face, une colonne en pierre pour rappeler les circonstances du Nirvana de Jou-laï. Elle porte, il est vrai, une inscription, mais on n'y a pas écrit le jour ni le mois de cet évènement.' (p. 334).

From this it is clear that either the Sramanas of Huen Thsang's time could not, or did not find it convenient to read the ancient inscriptions of Asoka.

The new text rectifies the imperfect identification afforded by the former version in regard to the succession to the throne of Kanauj: it now seems that Rájavarddhana was the monarch slain by Sasánka (p. 248); and that Harsha-varddhana 'accepta alors l'héritage de la royauté, se désigna lui-même par le nom de prince royal (Koumâra-râdja) et prit le titre de Cîlâditya.' (p. 251).]

contre lui, mit son armée en déroute, et le tua sur le territoire de Korour, situé entre Moultan et le château de Louny. Cette époque devint célèbre, à cause de la joie que les peuples ressentirent de la mort de Saca, et on la choisit pour ère principalement chez les astronomes.

و اما تاريخ بلب وهو صاحب مدينة بلبه وهي جنوبية عن مدينة انهلواره بقريب من ثلثين جوزن فان اوله متاخر عن تاريخ

شق بمائتي واحدي واربعين سنة و مستعملوه يضعون شككال الخمسة فبقي تاريخ بلب وينقصون منه مجموع مكعب السنة و مربع و خبره آت في موضعه و اما كوبت كال فكان كما قيل قوما اشرارا اقوياء فلما انقرضوا ارخ بهم و كان بلب كان اخيرهم فان اول تاريخهم ايضا متاخر عن شككال ۲۴۱

Ballaba, qui a donné aussi son nom à une ère, était prince de la ville de Ballaba, au midi de Anhalouara, à environ trente yodjanas de distance. L'ère de Ballaba est postérieure à celle de Saca de, 241 ans. Pour s'en servir, on pose l'ère de Saca, et l'on en ôte à la fois le cube de 6 (216) et le carré de 5 (25). Ce qui reste est l'ère de Ballaba. Il sera question de cette ère en son lieu. Quant au Goupta-kâla (ère des Gouptas), ou entend par le mot Goupta des gens qui, dit-on, étaient méchants et puissants; et l'ère qui porte leur nom est l'époque de leur extermination. Apparemment, Ballaba suivit immédiatement les Gouptas; car l'ère des Gouptas commence aussi l'an 241 de l'ère de Saca. . . . . D'après cela, en s'en tenant à l'an 400 de l'ère de Zezderdjed, on se trouve sous l'année 1488 de l'ère de Sri-Harscha, l'an 1088 de l'ère de Vikramaditya, l'an 953 de l'ère de Saca, l'an 712 de l'ère de Ballaba et celle des Gouptas. .

Déjà je me suis excusé sur l'imperfection de ce qui est dit ici, et j'ai averti que les résultats que je présente offraient quelque incertitude, vu les nombres qui excèdent celui de cent. Je ferai remarquer de plus que j'ai vu les Indiens, lorsqu'ils veulent marquer l'année de la prise de Soumenat (par Mahmoud le Ghaznévide), évènement qui eut lieu l'an 416 de l'hégire (= Janvier 1026 de J.C.), et l'an 947 de l'ère de Saca, je les ai vus écrire 242; puis, au-dessous 606; puis encore au-dessous, 99; enfin additionner le tout ensemble; ce qui donne l'ère de Saca. Ou peut induire de là que le nombre 242 indique les années qui précèdent l'époque où les Indiens commencèrent à se servir d'un cycle de cent, et que cet usage commença avec l'ère des Gouptas. D'aprés cela, le nombre 606 indiquerait les samvatsaras de cent complets, ce qui porterait chaque samvatsara à 101. Quant au nombre 99, ce seraient les années qui se sont écoulées du samvatsara non encore révolu. C'est ce qui est en effet. J'ai trouvé la confirmation et l'éclaircissement de cela dans les tables astronomiques de Durlab le moultanien; on y lit; 'Ecris 848 et ajoute le Loka-kâla, c'est-à-dire le comput du vulgaire; le produit marquera l'année de l'ère de Saca.' En effet, si nous écrivons l'année de l'ère de Saca qui correspond à l'année actuelle, et qui est l'année 953, et que nous retranchions de ce nombre la quantité 848, il restera 105 pour le Loka-kâla, et l'année de la ruine de Soumenat tombera sur le nombre 98.' Journal Asiatique, 4-me. série, tom. iv. (1844).

Having exhausted the somewhat voluminous but inconclusive documentary evidence bearing on the domination of the Guptas, I will briefly recapitulate the various arguments advanced for the determination of their era, which is probably of more real importance towards the justification of the subordinate periods of Indian history than any other epoch in the unrecorded annals of that land.

To Prinsep, it will be seen, is due the credit of having first proved "the Indo-Scythic paternity of the Kanauj [Gupta] coinage," (Art. x., Nov., 1835,) which he more explicitly developed in his paper on "The Transition from the Mithraic or Indo-Scythian coinage to the Hindú series,” (Art. xiii., Oct., 1836.)

Prof. Wilson, in his 'Ariana Antiqua,' (p. 418,) concurred in these identifications, giving even greater emphasis to the value of the 'evidence that the coins of the Gupta princes succeeded immediately to those of the Mithraic princes.'

Prinsep's conclusions with regard to the absolute date of these coins were modified, from time to time, with the advance of his knowledge, the details of which may be gathered from the essays here reprinted. Prof. Wilson, writing in 1840, avoided the expression of any very decisive opinion as to the epoch to which these issues should be attributed; and, though he was disposed to restrict the most modern period... to the 7th or 8th century,' he was prepared, on the other hand, to admit a possible antiquity reaching to 'the 2nd or 3rd century of our era.' ('Ariana Antiqua' pp. 417-8-9).

In the year 1848, having occasion to investigate the probable date of the Sáh kings of Saurashtra ('Jour. Roy. As. Soc.', xii., 1.), I was led to advert to that of the Guptas, which necessarily bore an important relation to the period of a race to whose dominions the latter dynasty succeeded. In considering this collateral branch of my subject, I was led to conclude that the dates occurring in the Gupta inscriptions might with propriety be referred to the Saka Samvat, and that the San 93' of the

1

Chandra Gupta record at Sanchi (suprâ cit.) corresponded therefore with 172 A.D.: a decision which has since been accepted by Prof. Lassen ('Ind. Alt.', ii. 939.) The chief groundwork of my deductions consisted, however, in the passage of Albirúni just quoted (p. 269), which, apart from the critical difficulties of the original text, seemed fully to authorise an inference that the Guptas preceded the Valabhis, and, moreover, as was therein shewn, and elsewhere proved, that the Valabhi æra reckoned from 319 A.D. Objection might be taken to my placing too much reliance upon the statement of an oriental writer, a foreigner in the country of whose history he was treating: but the author was clearly no superficial observer, and the statement itself fell so well into the fitness of things, and so nicely accorded with other indications bearing on the question, that I accepted it without hesitation, even as I now confess my faith in its verity to be rather confirmed than shaken by subsequent investigations and the new evidence that time has brought to light.

Prof. Lassen in reviewing the subject in 1852 ('Ind Alt.' ii.)

[I do not allude to the fanciful rectification proposed by Major Cunningham (Bhilsa Topes.' p. 139), but to a modification of meaning one portion of the original text will bear, and which recommends itself to our English scholars in preference to M. Reinaud's version. I am bound to add, however, that I have consulted M. Reinaud on the merits of this new translation, and that he adheres to his original rendering!

TRANSLATION: Again, the era of Bulub, who was Lord (or founder) of the city of Balabah, which lies to the south of the city of Anhalwarah, about 30 jozans. And the beginning (first) of this era dates 241 years after the Shaka era. And those who use it, take the date of the Shaka era, and deduct from it the sum of the cube of six plus the square of five, and the remainder is the Balab year, and the mention of it will come in its place. Again, the Kúbat Kál (Gupta era), that was, as is said, a wicked and powerful family; when it ceased it was dated from, and as it were (it would seem that), Balab was the last of them, for the first of their era also is 241 years after the Shaka Kál.'

It will be seen that the new translation, if accepted, scarcely modifies the original inferences derived from the fact that the Gupta era is reckoned from the fall of the family whose name it bears. Whether Bhalaba was the last of their race, or an alien rising upon their ruin, imports but little in the determination of the dynastic epoch, which, under either aspect, is proved to have preceded A.D. 318-19. But the revised translation certainly gets rid of one inconsistency, and explains more plausibly how one and the same era came to be called by two distinct names; and instead of that cycle having to be referred to the rise of one family and the fall of the other, it is under this interpretation reasonably attributable to the extinction of the single dynasty in the person of its latest potentate.]

280

Mahendra Gupta and Nârâjana Gupta

The authority for these assignments rests chiefly on the date of 93, noticed above, which is referred to the first ChandraGupta; and subordinately on the titles given in the Allahábád inscription to the Daiva-putra shahi shahán-shahi,' who is hence identified with Ardeshír Bábak. The author, therefore, places the rise of the Guptas soon after the death of Vikramaditya in 155 A.D.1

Major Cunningham, in his work on the Bhilsa Topes, has also reviewed at some length, and with little commendation, my proposed scheme for the determination of the era of the Guptas. I have replied to his arguments in another place,2 and I need not now re-open the discussion further than to refer to his chronological table inserted at the foot of page 264.

Since my first paper on the subject appeared in the 'Jour. Roy. As. Soc.', however, an apparently insignificant piece of progress has contributed materially to enlarge our view of the general bearings of the question, and tends rather to necessitate a shortening up of the period over which the rule of the Gupta succession should be spread. I allude to the decipherment of the names of Kumára and Skanda Gupta on the small silver pieces (Art. XI., infrá), which precede those bearing the designation of Budha Gupta, whose own money again is closely imitated by Toramána, the monarch whose inscription has been already inserted

1 [' 'Ind. Alt.', pp. 752, 937, 938, 939, 940, 942, 951, 961, 987, etc.]

2 [Jour. As. Soc. Beng.', v. xxiv., 371. (1855.)]